you know, it piss me off when i cant see my test site due to disk problem with the server. then i realize that i am on a different timezone.. stupid me..

good thing i read this thread before.. and yes this is helpfull..

***************

anyway.. i dont care now if the disk will lock or not, i am doing my test now locally on my pc and by the time i finish the test site.. the new accounts have been moved completely..

though on the lost files and data.. i lost some data and files..


seems that there must be a warning to be sent or a timeframe for a non safe updates of file..

------------------------------------

my files was gone, around 7 of them and mySQL database have rolled back a few minutes [3-4mins] and lost some database tables..

the server seems to have a time delay before it locks the files and the SQL database.. on the part of the SQL database.. some rollbacks do occur..

no worries anyway on my side since it was a test site.. the other users may have been affected though and have symptoms like mine.. by the way.. i am on smarty engine and part of my templates are on the database.. this may also contribute..

On our important production sites, each one of us should have efficient backup procedures, in order to recover from system crashes. There is no real system which is guaranteed free of crash possibility, so we should always be ready to recover from a disaster.
As SI, backups is our job. When I forget backing up my own files, I cannot complain if they disappear, it's my responsability. By the way, of course, if Xisto is able to do some important migrations without any data loss, thank a lot, good job folks!

I figured something was going on. This is not a complaint but a funny story.

I have a program that runs updates from my Astahost account. Basically it reads a directory and if the files there are newer than on the client, the files are downloaded and installed. Last night I kept getting one file that was always downloading and I couldn’t figure out why. After spending 3 hours in my debugger I finally went to that directory and found something strange. The file modified date was November 18, 2006. It is now November 9. I guess I warped to the future and updated the file, lol. As I said this is not a complaint but more of a story of weird things that can happen when debugging.

You should read the other growing topic at astahost talking about the fourth dimension, and travelling along the time axis.

that time warp made my compiled directory clean..

i have templates generated 7 days in the future which my moded smarty treats as invalid and keeps on trying to wipe it out..

Well if it is a problem you are experiencing it should be in a separate topic so not to spam the forum.

This explains a lot.

I had assumed that the server issues were due to either a down server, causing Astahost to cram are accounts on less servers for maintenance or that the number of free hosting accounts has just grown so large that the servers can't handle it.

Now that I see this thread though I completely understand what happened.

If I may make a small suggestion, thats if a system similar is not already in place.

These problems were easily forseen and it would have been nice to recieve some message/warning that the servers might be experiencing problems due to this kind of stuff.

That way we would not freak out when the servers are going down.
You should have seen my face when I realized all my files were permnantly read-only no matter how I applied CHMOD. I was completely baffled, after a few days of it I was almost ready to buy a hosting account sooner then I had planned.

Anyways it would be nice if we could be warned of future possible problems like this. Another example was the CPANEL update, I had no idea that cpanel was updated until I realized my php mail function stopped working.

I completely understand if you prefer to keep this information confidential, because they are your servers and you are providing us free accounts. If you already have a section like this well you can reach through your monitor and slap me in the face for not paying attention.

not a problem here.. a funny side effect..
it only just a matter of 1 char in my config this prevent the loop and it only happen 1 hour a day..

**********

and it makes me think what is going on.. so far.. i am not the only one who got the timewarp..

@BitShift

same point here.. though no damage on my side.. it helps anyway to warn..
i face the computer for 3 long hours trying to figure out why cant my smarty generate templates.. and 2 days later i stumble to this forum thread..

anyway.. the effect may have not been forseen.. i use to have this kind of updates on my past works.. where 300 computers checks my private mySQL server and downloads the binaries stored in blob fields..

5 months later.. 15 machines have been added and the whole update system crash mySQL server rendering some clients to have corrupted DLL's.. the problem was not the machine, it is the new network router with a lan cable connected to a network router which is also have a spare lan cable connected back to the other router.. result.. massive loopback that crash the network matrix overloading the mySQL server..

i guess something similar to this have happen considering the volume of data to build and transfer plus the server restart that needs to be done.. sometimes locking the filesystem is way much better than rendering your uploads in the middle while the server is being restarted..

Eh, the weird thing is that its still a read-only server for me. I don't know what's going on, but every time I try to upload something, I always get an error message saying it's read-only. Perhaps it stops me from uploading if I upload too many files at a time? *not sure what's going on*

Yeah, this process is still in effect from what I can tell. A feww accounts are transfered for the day, the server cleans up its configuration and reboots.

This usually happens between 0000 GMT and 0500 GMT. Which is between 8:00PM EST and 1:00AM EST. But only lasts a short time. This seems to be currently effecting the server as I type!

I don't think the outages will continue much longer.

vujsa